[2197] in linux-scsi channel archive
Exabyte problems
daemon@ATHENA.MIT.EDU (Tethys)
Thu Jul 24 08:48:31 1997
To: linux-scsi@vger.rutgers.edu
Date: Thu, 24 Jul 1997 12:28:20 +0100
From: Tethys <tethys@ml.com>
I've recently got an exabyte tape drive that I'm trying to get working.
It's recognised by the system, and I can rewind it, offline it and so
on with mt. However, when I try to read from or write to the tape, it
hangs and the process goes into an uninterruptable sleep. If I strace
the process, it's waiting on an uncompleted read (the very first one
that tried to access the tape). Presumably it'd be hung on a write
if I was writing to the tape.
Once it's got to that state, the only solution (nearly -- see later)
is to reboot. Even dropping down to single user mode kills all other
user processes, but leaves the one accessing the tape. After some 10
to 15 minutes hung, I occasionally get a SCSI timeout error message,
but the process stay hung.
I'm using an AdvanSys ISA 5140 SCSI adapter, with the following:
lazarus:~# cat /proc/scsi/scsi
Attached devices:
Host: scsi1 Channel: 00 Id: 02 Lun: 00
Vendor: Mountain Model: LCD-0001 1831416 Rev: 1.15
Type: Unknown ANSI SCSI revision: 01
Host: scsi1 Channel: 00 Id: 03 Lun: 00
Vendor: WangDAT Model: Model 1300 Rev: 02.4
Type: Sequential-Access ANSI SCSI revision: 02
Host: scsi1 Channel: 00 Id: 05 Lun: 00
Vendor: EXABYTE Model: EXB-8200 Rev: 252H
Type: Sequential-Access ANSI SCSI revision: 01
I have also had an HP 4020i CD writer, but disconnected it to
try and isolate the problem. Termination appears to be correct.
The DAT is an external unit that I added to see if it gave the same
problem (sadly, I only had one DAT tape, which the drive chewed
during testing, although it seemed to work before it did that).
The Mountain device is some kind of SCSI controller board in the
same box as the DAT, which, incidentally, appears to provoke the
following (apparently harmless) error message:
resize_dma_pool: unknown device type 31
Since I haven't yet got my SCSI disks installed, I've got SCSI support
compiled as a module (along with SCSI tape, CD and so on). Might this
be the problem? It seems to work quite happiliy with my CD writer with
modular SCSI support. I'll try recompiling this evening to see if it
makes a difference.
I later found out that unplugging the DAT drive caused enough SCSI
confusion to kill the hung process, so at least I no longer have to
reboot each time I want to try something new. Once the process is
dead, I can unload the advansys module, reconnect the DAT, reload
the module and carry on.
Any advice or suggestions would be greatly appreciated.
Tet
--
``There appear to be few if any technical reasons to move from UNIX to
Windows NT. The performance of Linux exceeds that of NT 4.0 and Linux
appears to be more reliable.'' -- David Korn, AT&T, author of ksh
--------------------+--------------+----------------------------------------
tethys@ml.com | Micro$oft: | Linux, the choice of a GNU generation.
tet@astradyne.co.uk | Just say no! | See http://www.uk.linux.org for details
